Anthropic Secures Multi-Gigawatt Compute Expansion with Google and Broadcom, Surpasses $30B Run-Rate
Anthropic has significantly accelerated the AI arms race, securing multiple gigawatts of TPU capacity starting in 2027 while revealing an unprecedented $30 billion annualized revenue run-rate.
The News
Anthropic has fundamentally altered the financial and infrastructural math of the AI industry. In a bombshell announcement, the company confirmed an expanded partnership with Google Cloud and Broadcom to secure "multiple gigawatts" of next-generation TPU capacity, slated to come online starting in 2027. This physical infrastructure expansion is staggering in scale, representing one of the largest single compute commitments in the history of the sector.
More critically, Anthropic CFO Krishna Rao revealed that the company's annualized run-rate revenue has skyrocketed to over $30 billion in 2026—a monumental leap from the $9 billion reported at the end of 2025. The company disclosed that over 1,000 enterprise customers are now spending in excess of $1 million annually on Claude, doubling that specific customer tier in less than two months.
The OPTYX Analysis
This is a watershed moment that redefines the competitive hierarchy of frontier AI. A $30 billion run-rate completely shatters the narrative that Anthropic is merely a safety-conscious alternative to OpenAI; they are operating at a commercial velocity that challenges the very top of the market. The specific mention of "multiple gigawatts" of power is equally revealing. Compute is no longer measured purely in chip allocation, but in raw energy draw. By locking in this infrastructure with Google and Broadcom out to 2027, Anthropic is securing the physical runway required to train Claude generations far beyond the current frontier.
This move also deepens the geopolitical and domestic focus of AI infrastructure, with Anthropic emphasizing that the vast majority of this new compute will be sited in the United States, building upon their previous $50 billion commitment to American infrastructure. It forces competitors to either match this capital expenditure or risk falling out of the frontier model tier entirely.
AI Platforms Impact
For enterprise leaders and technical architects, Anthropic’s explosive growth and secured compute pipeline make Claude an indispensable pillar for any multi-model strategy. The sheer volume of enterprise adoption—evidenced by the surge in million-dollar-plus contracts—suggests that Claude is winning the battle for highly complex, secure, and reliable enterprise workloads. Brands must stop treating AI platform selection as a speculative experiment and start architecting for long-term lock-in with platforms that have the literal power grid capacity to support future scaling. If your AI vendor cannot guarantee gigawatt-scale compute by 2027, they may not be able to serve your operational needs when the next generation of models arrives.